Search

BRAFCAPSTAN CRC is a European, retrospective, multicenter, observational study evaluating real-world treatment practices for patients with BRAFIn total, 255 patients (median age 66.0 years; 58.4% female) with BRAFThis study is, to date, the largest real-world analysis of patients with BRAF

Three posture asymmetries and their connections were assessed by a statistical study as rigorous as possible on the subjects of a sample from a population considered as normal after negative criteria. The comparison of probabilities observed with random probabilities shows that there is less than one chance in a thousand to make an error by rejecting the hypothesis that the asymmetries of the orthostatic posture are random. The definition of the limits of these asymmetries thus fits in the search for the conditions of a clinical examination of the posture.

Introduction/Background* Standard therapy for advanced ovarian cancer (OC) includes radical debulking surgery followed by first-line platinum-based chemotherapy, although neo-adjuvant chemotherapy can be used. Most women with newly diagnosed advanced OC relapse within 3 years of standard treatment. In SOLO1, patients with advanced OC and a BRCA1 and/or BRCA2 mutation (BRCAm), in complete or partial response following first-line platinum-based chemotherapy, received maintenance olaparib or placebo for up to 2 years or until progression. At 5-year follow-up, median progression-free survival (PFS) was 56 months with olaparib vs 14 months with placebo (hazard ratio 0.33; 95% confidence interval 0.25–0.43). The OVAL-1 study will provide evidence on real-world effectiveness of olaparib in patients with BRCAm advanced OC treated in the first-line maintenance setting in France, Italy and the UK. Methodology Retrospective cohort, pan-European multicentre observational study with data abstracted from medical records at several time points until ≥3 years after first olaparib dose (index date). Eligible patients are adult females with a BRCAm and advanced (FIGO stage III/IV) OC, who received their first olaparib (tablet) dose between January 2019 and June 2020 in the first-line maintenance setting via Early Access Programmes (Italy, UK), Temporary Use Authorization (France) or reimbursement following regulatory approval (Italy, UK). Main study endpoint is real-world PFS. Secondary endpoints include overall survival and response rates. The study will also describe surrogate measures of response and tolerability, including time to discontinuation, dose modifications (with reasons) and time to first and second subsequent treatment. Outcomes will be described by key subgroup status pre-index, including performance status, FIGO stage, BRCAm status, debulking surgery outcome and clinical response to chemotherapy. The study aims to include 350 patients. Retrospective data collection began in December 2020 and is planned to end by Q3 2023. As of April 2021, 69 patients have participated.

Resume Introduction Le benefice des inhibiteurs de la tyrosine kinase specifiques de l’EGFR, chez des patients sans mutation de l’EGFR, est controverse. Methodes La survie de patients traites par erlotinib en 2e ou 3e ligne pour un cancer bronchique non a petites cellules (CBNPC) sans mutation activatrice de l’EGFR a ete evaluee a partir de donnees en vie reelle issues d’une cohorte prospective, nationale, multicentrique, non-interventionnelle. Resultats Au total, 274 patients ont ete analyses, dont 185 (68 %) traites par erlotinib et 89 (32 %) exclusivement traites par des soins de support. La mediane de survie globale etait de 4,2 mois (IC95 % [3,5 ; 5,4]) avec erlotinib et 1,3 mois (IC95 % [1,0 ; 1,8]) avec les soins de support. A 3, 6, et 12 mois, le taux de survie etait de 62 %, 37 %, et 17 %, respectivement, avec erlotinib, contre 20 %, 8 %, et 3 %, avec les soins de support. Les facteurs pronostiques d’une survie globale plus longue etaient la presence d’un adenocarcinome et une 1re ligne de chimiotherapie incluant des taxanes, du pemetrexed ou de la vinorelbine (p Conclusion L’erlotinib reste une option therapeutique pour les formes localement avancees ou metastatiques du CBNPC apres echec d’au moins une ligne de chimiotherapie, chez des patients ineligibles a une chimiotherapie.

Background Patients with metastatic non-small-cell lung cancer (NSCLC) who survive more than 2 years are considered long-term survivors (LTSs). The present study examined factors associated with long-term survival and collected information for future comparison. Methods Clinical, molecular, and therapeutic data were collected from patients followed for primary stage IV (7th TNM classification) NSCLC within 2 years from diagnosis in the respiratory medicine departments of 53 French non-teaching hospitals. LTS and non-LTS records were compared. Factors associated with long-term survival were examined by univariate and multivariate analyses using logistic regression models. Results Vital status at least 2 years after diagnosis was known for 1977 stage IV NSCLC patients; 220 (11.1%) were LTSs. On multivariate analysis, independent positive factors comprised: TTF-1(+) immunochemistry, EGFR-mutation, surgery, rescue radiotherapy, and targeted therapy. Independent negative factors comprised: prediagnosis weight loss > 5 kg, ECOG performance status > 1, and primary radiotherapy. Conclusions Molecular biology and targeted therapy were decisive for long-term survival. With their development and their widespread implementation in clinical practice, the percentage of LTSs is expected to grow. Factors determining long-term survival found in this study should be taken into account when considering treatment options for patients with stage IV NSCLC.

The benefit of tyrosine kinase inhibitors for patients with an EGFR wild-type non-small cell lung cancer (NSCLC) remains controversial.The survival of patients with an EGFR wild-type NSCLC who received second- or third-line erlotinib treatment was assessed using real-life data that had been collected in a prospective, national, multicenter, non-interventional cohort study.Data from 274 patients were analysed, 185 (68%) treated with erlotinib and 89 (32%) treated with supportive care only. The median overall survival was 4.2months (95% CI [3.5; 5.4]) with erlotinib, and 1.3months (95% CI [1.0; 1.8]) with supportive care. Survival rate at 3, 6, and 12months was 62%, 37%, and 17%, respectively, with erlotinib, versus 20%, 8%, et 3%, with exclusive supportive care. Significant predictive factors for longer overall survival were the presence of adenocarcinoma, and use of 1st line chemotherapy including either taxanes, pemetrexed or vinorelbine (P0.05).Erlotinib remains a valuable therapeutic option to treat inoperable locally advanced or metastatic NSCLC after failure of at least one prior chemotherapy regimen in fragile patients who are not eligible for chemotherapy.

Introduction Le nivolumab a demontre un benefice en survie globale (SG) chez les patients pretraites avec un cancer bronchopulmonaire non a petites cellules (CBNPC) avance dans deux etudes randomisees de phase III. Methodes EVIDENS est une etude observationnelle francaise multicentrique longitudinale ayant inclus entre decembre 2016 et novembre 2017 des patients atteints d’un cancer bronchopulmonaire et traites par nivolumab. Nous rapportons ici les resultats d’une analyse intermediaire avec une duree de suivi minimum de 8 mois. La SG et la survie sans progression (SSP) ont ete estimees a l’aide de la methode de Kaplan–Meier. Resultats L’etude EVIDENS a inclus 1394 patients atteints d’un CBNPC avance et ayant recu au moins une injection de nivolumab. Les sous-types histologiques epidermoides et non-epidermoides concernaient respectivement 31,1 % et 68,9 % des patients. Les caracteristiques des patients a l’inclusion etaient representatives de la population standard de patients atteints d’un CBNPC avance : 66,0 ans d’âge median, 69,2 % d’hommes, 89,6 % de fumeurs actuels ou anciens fumeurs et 83,2 % ayant un PS 0–1. Des metastases cerebrales et des maladies auto-immunes actives ont ete respectivement decrites chez 20,0 % et 2,9 % des patients a l’initiation du traitement. L’expression de PD-L1 a ete evaluee chez 18,9 % des patients et 63,6 % d’entre eux avaient une tumeur consideree comme positive (≥ 1 % des cellules tumorales ou immunitaires positives). La recherche d’une mutation de l’EGFR a ete effectuee chez 91 % des patients presentant une tumeur non-epidermoide, et rapportee chez 5,6 % d’entre eux. Nivolumab a ete administre en 2e et 3e ligne ou plus respectivement pour 74,0 % et 25,4 % des patients. Le nombre median d’injections etait de 5. La mediane de SSP etait de 2,8 mois et la mediane de SG estimee a 10 mois. A 6 mois, le taux de SSP etait de 30,5 %, le taux de SG de 65,3 % et le meilleur taux de reponse objective de 16,5 %. Des evenements indesirables ont ete decrits chez 30,9 % des patients traites et 7,0 % ont presente au moins un grade 3 ou 4. Les evenements indesirables d’origine immunologique ont ete decrits chez 12,4 % des patients traites et 1,5 % ont presente au moins un grade 3 ou 4. Les arrets de traitement pour cause de toxicite ont ete rapportes chez 4,6 % des patients. Conclusion Les resultats preliminaires d’EVIDENS confirment en vie reelle le profil de tolerance et l’efficacite du nivolumab administre dans les centres francais chez des patients presentant un cancer bronchopulmonaire.

Introduction En oncologie, l’indice fonctionnel ECOG (PS), echelle de mesure de l’etat de sante general des patients, est un facteur pronostique systematiquement pris en compte dans les essais cliniques. Dans la pratique medicale courante, le PS est utilise par le medecin pour adapter la strategie therapeutique mais non systematiquement documente dans le dossier medical du patient. Le PS ne pouvant etre reporte retroactivement, la construction d’un substitut du PS permettrait de prendre en compte ce facteur dans les analyses sur donnees longitudinales retrospectives en vie reelle. L’objectif de ce travail etait de construire un substitut du PS a partir d’informations disponibles (âge, comorbidites, constantes biologiques) et decrire la performance du modele de prediction. Methodes A partir de la plateforme epidemio-strategie et medico-economique (ESME) des CLCCs sur le cancer du sein metastatique (CSM), les donnees de 865 patients traites dans quatre centres pour leur maladie metastatique a partir de 2014 ont ete extraites. Un recueil retrospectif a ete a ete organise concernant les comorbidites de Charslon et des valeurs biologiques (hematologie et biochimie). Un modele de regression logistique a ete utilise pour predire le substitut du PS. Les variables utilisees pour predire le PS etaient : l’âge, le statut menopausique, le score de Charslon, les transaminases, la bilirubine, la creatinine, les LDH, l’hemoglobine et les lymphocytes. Le PS a ete evalue selon deux seuils (PS > 0 et PS > 1). La performance du modele a ete utilisee par la methode de l’aire sous la courbe de ROC. Resultats Une mesure du PS au diagnostic de la maladie metastatique etait documentee chez 28 % seulement des 865 patients identifies. Les 619 patients sans mesure du PS ont ete exclus de l’analyse (316 patients sans aucune mesure du PS disponible dans le dossier medical et 303 patients avec une mesure du PS non disponible au diagnostic metastatique). Les patients exclus etaient un peu plus âges (63,2 versus 60,0, p 0 pour lequel l’âge (OR = 1,01 ; IC95 % [1,02–1,09]) et la bilirubine (OR = 8,01 ; [1,31–50,40]) etaient significativement associes. Les memes variables ont ete suggerees a partir d’une cohorte de validation externe comprenant 164 patients malgre une performance peu informative (AUC Conclusion Ce premier travail sur la prediction d’un substitut du PS suggere l’âge et un marqueur hepatique comme variables candidates pour integrer le modele de prediction. Malgre son utilisation courante dans le domaine de la prediction, le score de Charlson n’a pas ete retenu dans le modele final. L’evaluation de la performance du modele etait moyennement informative. Une etude prospective avec un effectif plus important et au diagnostic du cancer du sein metastatique, une mesure systematique du PS, l’administration d’un questionnaire de qualite de vie et la collecte de constantes biologiques permettrait de confirmer ces resultats.

The objective of the ESCAP-2011-CPHG cohort study was to perform a real-life analysis of therapeutic strategies used during the first 2years of follow-up after a diagnosis of primary lung cancer. This paper presents the study and its first results in non-small-cell lung cancer (NSCLC).Pulmonologists in the respiratory disease departments of 53 general hospitals consecutively included all patients aged 18years and over with lung cancer newly diagnosed in 2010.Of the 3943 patients included, 3418 (mean age: 65.4 years; male: 76%; never smokers: 12%) had NSCLC (adenocarcinoma: 53%; stages 0-II, IIIA, IIIB and IV: 18, 14, 9 and 59%, respectively). Mean follow-up was 13.2 (SD: 10.1) months; mean number of strategies implemented was 2 (SD: 1.3). Overall, 62% of patients had chemotherapy in the first strategy (74% in the second strategy); the rate of chemotherapy alone increased from 6 to 56% with cancer stage.ESCAP-2011-CPHG opens the way to many possible analyses of the therapeutic strategies currently implemented in French hospitals, comparing strategies, survival or patient characteristics.

The French college of general hospital respiratory physicians (CPHG) has conducted 10 years apart two prospective observational studies to assess changes in the primary lung cancer epidemiology and outcomes, including 1-year mortality.In 2000 and 2010, all volunteer adult patients followed in the respiratory department of general hospitals participating in the study were consecutively included if their lung cancer was histologically or cytologically diagnosed between 01 January and 31 December (sample date). Their vital status at least 1 year after inclusion and date of death (if applicable) were collected.Respectively, 5667 and 7051 patients were included in the study in 2000 and 2010 and vital status of 5441 (96.0%) and 6981 (99%) patients known. One-year mortality rate was 61.8% in 2000 and 56.4% in 2010 (P0.0001). Mortality rate significantly decreased from 2000 to 2010 in non-small-cell lung cancer (60.7% vs. 55.2%; P0.0001) but not in small-cell lung cancer (66.9% vs. 64.2%; P=0.22). The year of diagnosis was an independent risk factor of mortality (OR=0.84; 95% CI: 0.77-0.91; P0.0001).Although it remains low (43.6% in 2010), life expectancy at 1 year for patients with lung cancer has improved in 10 years. Five-year results are expected to show whether this improvement is maintained or not over time.

Introduction Le nivolumab est un anti-PD1 disponible en France depuis janvier 2015, dans le cadre d’une autorisation temporaire d’utilisation (ATU) puis de l’autorisation de mise sur la marche (AMM), a partir de la 2e ligne de traitement du CBNPC localement avance ou metastatique. A ce jour, peu de donnees sont disponibles en vie reelle. Methodes EVIDENS est une etude observationnelle francaise multicentrique longitudinale qui vise a inclure pres de 1300 patients traites par nivolumab pour un CBNPC avance. L’objectif principal est de decrire les caracteristiques sociodemographiques et cliniques des patients a l’initiation du nivolumab et d’estimer la survie globale a 3 ans dans la population globale et par sous-type histologique. Les objectifs secondaires sont d’evaluer a 1, 2 et 3 ans la survie globale, la survie sans progression, le taux de reponse objective, l’evolution de la qualite de vie ainsi que l’incidence, la gravite et la prise en charge des effets indesirables. L’evaluation de la reponse est faite par l’investigateur. Resultats Au 30 juin 2017, 471 patients ont ete inclus. Le Tableau 1 resume leurs caracteristiques. Les patients sont majoritairement des hommes (69,6 %), âges de moins de 75 ans (84,4 %) atteints d’un CBNPC non epidermoide (67,9 %). Les fumeurs actifs ou sevres representent 95,4 % des cancers epidermoides et 89,4 % des cancers non epidermoides. A l’initiation du nivolumab, 83,4 % des patients ont un statut ECOG ≤ 1, 91,7 % ont un CBNPC de stade IV, 4,2 % presentent une mutation EGFR et 1,1 % un rearrangement ALK. Pres de 24 % des patients de stade IV presentent des metastases cerebrales, majoritairement asymptomatiques (83,8 %). Le nivolumab a ete administre en 2e et en 3e ligne pour respectivement 75,2 % et 16,8 % des patients. L’expression de PD-L1, evaluee par immunohistochimie chez 17,8 % des patients, a ete retrouvee chez 0 % des cancers epidermoides testes (n = 5) et 64,8 % des cancers non epidermoides testes (n = 54) avec une mediane a 35 % de cellules tumorales marquees [1–100]. A l’inclusion, la qualite de vie des patients, mesuree par l’echelle EQ-5D, etait sans difference selon le type histologique. Conclusion L’analyse preliminaire d’EVIDENS apporte une description precise de la population traitee en conditions de vie reelle par nivolumab pour un CBNPC avance en France. Les prochaines analyses permettront d’evaluer l’efficacite et la tolerance de ce traitement, ainsi que l’evolution de la qualite de vie des patients.

Background: Circadian clocks involve 15 specific genes, which control cell cycle, apoptosis, DNA repair and metabolism. Circadian disruption is associated with both an increased risk of breast cancer, and poor survival in patients (pts) with metastatic cancer (IARC#98; Innominato et al. Cancer Res 2009). Objective: to provide first evidence of clinical relevance of clock genes expression for the therapeutic sensitivity of primary breast cancer. Method: Affymetrix Hgu133plus2 microarrays data were derived from 189 primary breast cancers. All the pts had non metastatic disease and were registered in multicenter Phase II trial of neoadjuvant chemotherapy with epirubicin-cyclophosphamide (4 courses) then docetaxel ± herceptin (4 courses). In this study, negative hormonal receptor status and limited tumor size predicted for pathological complete response rate (pCR) (Pierga JY, et al Breast Cancer Res Treat. 2010).The expression data of 15 core clock genes, within the whole tumor transcriptome, were related both to tumor mitotic index (MI) and estrogen receptors (ESR1), and to chemotherapy-induced pCR. Statistical validation involved T-test statistics, with corrections for multiple testing. Results: In the tumor biopsy obtained before treatment onset, both Per2 and Cry2 expressions correlated positively with ESR1 (r≥0.43; P Clock gene expression and pCR Conclusion: Low expressions of three core genes in the negative loop of the molecular circadian clock strongly predicted for the induction of a pathologic complete response of primary breast cancer with neoadjuvant chemotherapy. While circadian clock disruption seems to constitute a poor prognostic factor in cancer patients, our data suggest that the downregulation of core circadian genes in tumors significantly enhances susceptibility to chemotherapy. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-09-26.

Resume Selon une approche qualitative basee sur le temoignage et la narration, un groupe de reflexion et de recherche au sein de l’Espace ethique Assistance publique–Hopitaux de Paris a explore la question de « l’erreur medicale », ses circonstances et ses enjeux dans la relation entre les soignants, les parents et l’enfant. Plusieurs points forts emergent de ce travail : certaines circonstances sont susceptibles d’augmenter le risque de survenue d’une erreur : contexte hospitalier et societal, manque d’ecoute, climat de defiance, defaut de procedures de controle, situations de souffrance extreme… L’erreur est souvent le resultat d’une accumulation de defaillances successives ; il est possible de traverser l’erreur ensemble : parents, enfant et soignants ; cela depend de la qualite de la relation de confiance etablie depuis le debut de la maladie, mais aussi de la facon dont l’erreur est reconnue, communiquee et geree et ce quelle qu’en soit la gravite ou les consequences pour l’enfant ; en effet les consequences seules ne determinent pas le vecu de l’erreur : meme sans consequence tragique, une erreur peut engendrer un traumatisme durable. A contrario, des consequences tragiques pouvant aller jusqu’a la mort n’entrainent pas necessairement la rupture de la confiance. Les maitres-mots sont vigilance et maintien en eveil permanent. C’est d’un questionnement ethique et collectif sans relâche au sein des equipes hospitalieres et de l’institution que depend la capacite des soignants a faire de l’erreur un moteur d’amelioration des pratiques de soins et a donner a l’enfant et a chacun des membres de sa famille la possibilite de poursuivre l’ecriture de leur histoire personnelle et familiale dans un contexte si exceptionnellement douloureux.

This prospective, observational study investigated the haematological response to darbepoetin alfa (DA) administered every three weeks for the treatment of anaemia. Response was also assessed according to baseline characteristics including iron, folate and vitamin B12 status.Anaemic adult patients with malignant non-myeloid cancer, starting or having already undergone chemotherapy received DA on day of inclusionand were followed up for up to 24 weeks. Concentration of haemoglobin (Hb), as well as iron, vitamin B12 and folate status where available, were recorded at inclusion, after a treatment period of 9 weeks and up to a maximum of 24 weeks or cessation of DA treatment, whichever was sooner.The main outcome measure assessed in this study was the percentage of patients reaching a Hb concentration of at least 11 g/dL at least once at any time during the study.A total of 2912 patients were included. The mean Hb concentration increased from 10.0 g/dL at inclusion to 11.4 g/dL at 9 weeks and 11.8 g/dL at 24 weeks. In 74.6% of patients the target Hb level of 11.0 g/dL or above was reached. After initiation of DA treatment, 9.5% of patients required a blood transfusion by week 9, and 5.6% thereafter. Vitamin B12 and folate status were unknown for 80.3% of patients and the iron status for 73.2% of patients. Compared with patients who remained untreated for vitamin B12 or folate deficiency, a higher percentage of patients with vitamin status within normal limits achieved the target Hb concentration. However, achievement of target Hb level appeared not to be affected by iron status.In this study, the mean Hb level increased in anaemic cancer patients treated with DA and the majority of patients achieved the target Hb level. In contrast to the recommendations of guidelines (EORTC) encouraging the measurement of iron and vitamin levels, the present study demonstrated that data were not routinely collected for these factors.

The effectiveness of three and six cycles of MOPP (chlormethine, vincristine, procarbazine and prednisolone) as initial chemotherapy in 108 patients with Hodgkin's disease at clinical stages II nA, IIB, IIIA and B was compared in a study of two groups, one of them treated in six cycles (October, 1972--December, 1976), the other in three cycles (January, 1977--October, 1979). The clinically complete remissions obtained by this schedule and findings at exploratory splenectomy were chosen as measures of effectiveness. The frequency of clinically complete remission was similar in both groups. In 96% (97% for the second group) of patients in whom the three (six) MOPP cycles of cytostatic treatment achieved clinically complete remission, exploratory splenectomy failed to reveal any infradiaphragmatic involvement. In patients in stage IIA and IIB one can, therefore, consider reducing primary cytostatic treatment to three MOPP cycles. Exploratory splenectomy and prophylactic infradiaphragmatic radiotherapy can be omitted in these patients, if one accepts the 5% risk of infradiaphragmatic involvement. Splenectomy can also be omitted in patients in stage III, but not infradiaphragmatic radiotherapy, including that no the spleen.

Objective: To assess the efficacy and safety of pentoxifylline, a US Food and Drug Administration–approved drug, in patients with ALS treated with riluzole. Methods: The authors conducted a double-blind, randomized, placebo-controlled, multicenter trial. Four hundred patients with probable or definite ALS and vital capacity less than 100% were randomly assigned to treatment with placebo or 1.2 g pentoxifylline daily. The primary outcome was death. Secondary outcomes were rates of deterioration of ALS Functional Rating Scale–Respiratory and muscle strength. The primary intention-to-treat analysis was the survival comparison of drug vs placebo, assessed before (log-rank test) and after adjustment (Cox model) for predefined prognostic factors. Results: At the end of the study, after 547 days of follow-up, 103 patients (51.7%) in the pentoxifylline group and 120 (59.7%) in the placebo group were alive (unadjusted risk 1.28, p = 0.107; adjusted risk 1.43, p = 0.02). In contrast, analysis of secondary outcome functional variables did not show the same negative effect of the drug. The most common adverse reactions were nausea, dysphagia, and flushing, all reversible after stopping the drug. Conclusions: Pentoxifylline is not beneficial in ALS and should be avoided in patients treated with riluzole. The discrepancy between survival and measures of functional changes urges caution in equating these end points in phase III trials, and suggests that both survival and function should be used in phase III trials.

This paper considers the utility of spline approximations of the hazard function for estimation of cure rates from survival data. With this approach, it is possible to assess dissimilar effects of various covariates on the probability of tumor cure and on the timing of the endpoint (death, tumor recurrence, etc.) under study. The usefulness of the approach is illustrated with an application to data on survival of patients treated for breast cancer.

We report hereby the results of the french multicentric randomized PEGASE 04 protocol established to evaluate the impact on survival of high-dose chemotherapy over conventional chemotherapy for MBC patients.

Retrospective study of local tolerability of a natural hydroxyapatite orbital implant wrapped with Vicryl(®) (polyglactin) mesh in patients undergoing enucleation.Complications were classified into four types according to their management: type 1 if no reoperation was required, type 2 if additional surgery without grafting was required, type 3 if an oral mucosal graft was performed (major dehiscence) and type 4 if the complication required removal of the implant.Seven hundred and four patients with a median follow-up of 44 months. Five hundred and three patients were enucleated as a primary procedure and 201 after failure of conservative management. The overall complication rate was 12.07% (85 patients) with 68 type 1 complications, nine type 2 complications, three type 3 complications and five type 4 complications (0.71%). A total of 17 patients (2.42%) required additional surgery. The use of chemotherapy or radiotherapy before or after surgery did not influence the results. In univariate analysis, the tolerability was better in children than in adults. With multivariate analysis, only the diameter of the implant was an independent risk factor for complications (P=0.001).Use of a Vicryl(®) mesh-wrapped natural hydroxyapatite orbital implant after enucleation is particularly well tolerated including the pediatric population. A compromise should be sought with an implant large enough for good cosmetic results but small enough to avoid complications.

The evaluation of cancer treatment efficiency is often based on the risk of local recurrence. When applied to the statistical analysis of tumor recurrence data, a pertinent parametric method has the following distinct advantages: (1) it allows a natural interpretation in terms of parameters bearing clear biological meaning, (2) it provides a prediction of the recurrence risk forward in time beyond the follow-up period, (3) it offers a means of estimating survival fraction (probability of tumor cure) from the time-to-recurrence observations. This paper discusses a stochastic model of tumor recurrence based on the consideration of biological processes of tumor latency within the random minima framework. A parametric family of distributions is obtained that allows for a survival fraction, thereby providing an estimate of the probability of tumor cure. When applying the model to data on breast cancer, we estimate the expected number of clonogens which give rise to early and late recurrences and their progression rate parameters. The prime object of our concern is discrimination between true recurrence and spontaneuos carcinogenesis on the basis of the temporal characteristics of the tumor latency. As evidenced by the data analysis, such a discrimination is feasible and allows to conclude that the contralateral breast cancer may be interpreted as a preexisting subclinical tumor at the time of treatment.

Purpose There is a causal association between Helicobacter pylori (Hp) gastric infection and the development of gastric MALT lymphoma. In contrast, the link between Hp gastric infection and the development of extra-gastric lymphoma has not been thoroughly investigated. We therefore studied the prevalence of gastric Hp infection at initial diagnosis of ophthalmologic and non-ophthalmologic extra-gastric lymphoma patients. Methods Three cohorts of patients were studied: a first one of 83 patients with OAL, a second one of 101 patients with extra-ophthalmologic extra-gastric lymphoma, and a third one of 156 control individuals (control) without malignant lymphoma. Gastric Hp infection was investigated by histopathological analysis and Hp-specific PCR assay on gastric biopsy tissue samples. Results We found gastric Hp infection in 37 OAL patients (45%), in 25 extra-ophthalmologic extra-gastric lymphoma cases (25%), and in 18 controls individuals (12%) (p

The College of General Hospital Respiratory Physicians (CPHG) is following up the KBP-2000-CPHG study, performed ten years ago, with a new observational epidemiological study of primary lung cancer.The study includes all new cases of primary lung cancer diagnosed on histology or cytology between 1 January and 31 December 2010 and managed by one of the general hospital pneumology departments participating in the study. The primary objective is to estimate five-year mortality and to analyze risk factors. Secondary objectives are to describe the characteristics of this patient population and their management, and to estimate one, four and five-year survival rates. These data will be compared to those of the KBP-2000-CPHG study. The four-year prognosis score developed in 2000 will also be assessed in the new study. Data are collected by standardized questionnaire with exhaustiveness control.One hundred and eight investigating centers have agreed to participate; 4000 to 5000 new cases of primary lung cancer should be collected and analyzed.The study will describe the characteristics of patients presenting with primary lung cancer in the participating pneumology departments during the year 2010, and their diagnostic and therapeutic management, and assess changes over the last ten years.

This study aimed to evaluate the potential impact of social inequalities on stage at diagnosis and long-term outcome of breast cancer patients attending the Institut Curie in Paris (France).The study population included 14,610 breast cancer patients diagnosed and treated in the Institut Curie between 1981 and 2001. The socioeconomic status was determined from district of residence, median income for town of residence corrected by the consumption unit and body mass index. Logistic regression models adjusted on socioeconomic factors were used to evaluate clinical and pathologic features at diagnosis. Overall survival and distant metastasis were analysed with log-rank tests and Cox proportional hazards regression models.Patients living in lower income districts were more likely to be diagnosed with breast tumors size greater than 20 mm (P=0.01). Residents of high-income urban areas (15,770 €) exhibited a significant overall survival and distant metastasis advantage (respectively HR=0.93 [0.86-0.99]; P=0.02 and HR=0.91 [0.85-0.98]; P=0.01). Breast cancer screening with mammography was independent of district of residence (P=0.61) or income (P=0.14). After adjusting for age at diagnosis and period, the risk of having breast cancer with unfavorable prognostic factors such as tumor size greater than 20 mm decreased with 1000 € increase in district income (OR=0.986 [0.98-0.99]; P0.001). Similarly, the risk of cancer death decreased for patients residing in districts with median income greater than 15,770 € (HR=0.92 [0.86-0.98]; P=0.01).Despite the limitations of the study (aggregate data used to assess socioeconomic status, non representative cohort of French women), we observed that poorer breast cancer prognosis with advanced disease diagnosis and increased risk of breast cancer mortality was related to low socioeconomic status.

Several risk estimation models for breast or ovarian cancers have been developed these last decades. All these models take into account the family history, with different levels of sophistication. Gail model was developed in 1989 taking into account the family history (0, 1 oror = 2 affected relatives) and several environmental factors. In 1990, Claus model was the first to integrate explicit assumptions about genetic effects, assuming a single gene dominantly inherited occurring with a low frequency in the population. BRCAPRO model, posterior to the identification of BRCA1 and BRCA2, assumes a restricted transmission with only these two dominantly inherited genes. BOADICEA model adds the effect of a polygenic component to the effect of BRCA1 and BRCA2 to explain the residual clustering of breast cancer. At last, IBIS model assumes a third dominantly inherited gene to explain this residual clustering. Moreover, this model incorporates environmental factors. We applied the Claus, BRCAPRO, BOADICEA and IBIS models to four clinical situations, corresponding to more or less heavy family histories, in order to study the consistency of the risk estimates. The three more recent models (BRCAPRO, BOADICEA and IBIS) gave the closer estimations. These estimates could be useful in clinical practice in front of complex analysis of breast and/or ovarian cancers family history.

Our objective was to study the prognostic value of stoutness in non-metastatic breast cancer, from a population of French women.We constituted a large observational cohort of patients followed since a first unilateral breast cancer without distant dissemination. Stoutness was assessed using the body mass index (BMI, in kg/m(2)) at the time of diagnosis. Patients were classified into the four main categories of BMI, defined according to the World Health Organization recommendations. The risk of prognosis events was analyzed according to the BMI categories. To this end, survival analyses were achieved.The patients having a BMI value of at least 25 kg/m(2) presented significantly higher risks of death and metastasis recurrences when they were compared to the patients having a normal value of BMI. The multivariate analyses found a modest increase of risk, about 10 to 20%, depending on the degree of fatness. It reached about 20 to 50% according to the univariate analyses. The obese patients (BMIor= 30 kg/m(2)) had an increase of 50% of the risk of second primary cancers, comparatively to the patients having a normal value of BMI. Regarding contralateral, nodal and local recurrences, the survival analyses did not achieve any significant relationship with stoutness.A poorer prognosis is observed when breast cancer patients have a value of BMI that matches at least with overweight. Contrary to the results of few recent surveys, underweight patients do not present a poorer prognosis than normal weight patients. Excess of weight represents a modifiable factor in order to improve female breast cancer prognosis.

In 2000 the College of Pulmonologists of General Hospitals undertook an epidemiological study (KBP-2000-CPHG) enrolling all new cases of histologically confirmed lung cancer managed in general hospitals. This paper reports the five year survival in these cases.Vital status was available for 5447 out of 5667 patients included in the original study. The effect of different prognostic factors on mortality was assessed.At 5 years 567 patients (10.4%) were still alive. Median survival for the 4880 (89.6%) deceased patients was 7 months. Univariate analysis identified age, smoking history, performance status, histological type and disease stage (TMN classification) as determinants of survival. For non-small cell lung cancer (n=4885) multivariate analysis identified five predictive factors for mortality - age, gender histological type, performance status and stage.Five year survival in lung cancer continues to be poor. As the risk factors for poor outcome at the time of diagnosis are not modifiable and pending the results of screening studies reduction in mortality must rest on primary prevention.

Introduction Plusieurs etudes ont evoque le lien entre obesite ou amaigrissement au diagnostic et survie lors de CBP. Le College des pneumologues des hopitaux generaux (CPHG) a realise une etude prospective observationnelle multicentrique decrivant les caracteristiques et le devenir des patients avec un CBP. Cette analyse post-hoc etudie le lien entre mortalite et IMC tout stade et histologie confondus ( Fig. 1 ). Methodes Sept mille cinquante et un patients adultes suivis pour un CBP diagnostique entre le 01/01/2010 et le 31/12/2010 dans l’un des 104 hopitaux generaux participant ont ete inclus et classes selon leur IMC en 4 groupes : patients denutris, IMC 2 si 2 si ≥ 70 ans ; de poids normal, IMC ≥ 18,5 kg/m 2 si 2 si ≥ 70 ans et 2 ; en surpoids, IMC ≥ 25 kg/m 2 ; ou obeses, IMC ≥ 30 kg/m 2 . Resultats Respectivement, 15 %, 45 %, 28 %, et 12 % des patients etaient denutris, de poids normal, en surpoids, et obeses. Parmi les patients denutris ayant recemment maigri, 31 %, 45 %, et 24 % avaient perdu 10 kg. La survie a 1 an etait moins bonne chez les patients denutris (32 % contre 43 %, 50 %, et 53 % ; p Fig. 1 ) en partie liee a un moins bon etat general (22 % de performance status de 3 ou 4 contre 11 %, 8 %, et 7 % ; p p p = 0,242), alors que le surpoids (HR = 0,92 [0,85 ; 0,99] ; p = 0,037) et l’obesite (HR = 0,9 [0,81 ; 1,01] ; p = 0,062) etaient des facteurs protecteurs. Conclusion Lors de CBP, la survie a 1 an est particulierement mauvaise chez les patients denutris. Le statut nutritionnel est une variable pronostique a integrer dans le schema therapeutique, necessitant une prise en charge precoce. Une analyse tenant compte de la perte de poids dans les 3 mois chez les patients denutris est en cours.

Studying the molecular stratification of breast carcinoma is a real challenge considering the extreme heterogeneity of these tumors. Many patients are now treated following recommendation established at several NIH and St Gallen consensus conferences. However a significant fraction of these breast cancer patients do not need adjuvant chemotherapies while other patients receive inefficacious therapies. High density gene expression arrays have been designed to attempt to establish expression profiles that could be used as prognostic indicators or as predictive markers for response to treatment. This review is intended to discuss the potential value of these new indicators, but also the current weaknesses of these new genomic and bioinformatic approaches. The combined analysis of transcriptomic and genomic alteration data from relatively large numbers of well annotated tumor specimens may offer an opportunity to overcome the current difficulties in validating recently published non overlapping gene lists as prognostic or therapeutic indicators. There is also hope for identifying and deciphering signal transduction pathways driving tumor progression with newly developed algorithms and semi quantitative parameters obtained in simplified in vitro or in vivo models for specific transduction pathways.

This study aimed to estimate survival at 5 years by localization, sex and stage of the patients who presented a new cancer in 1994 in the Ile-de-France area.A cohort study began in 1994 by an exhaustive collection of the incidental cancers notified in the Ile-de-France area at the health insurance (27,080 patients). A stratified random sample based on tumor localization was followed at 1 year, 3 years and 5 years. The analysis of the observed survival was carried out according to Kaplan-Meier method. Relative survival was calculated according to Ederer II method.The follow-up sample concerned 4,166 patients. For all cancers, relative survival at 5 years was 65% for the women and 51% for the men. Relative survival rate at 5 years was 82% for the women with a breast cancer (98% for the patients in stage I) and 66% for those with a cancer of the cervix. Relative survival at 5 years for colonic cancer was 67% for men and 54% for women. For lung cancer, the relative survival rate fell from 47% for patients in stage I to 5% for those in stage IV.Our study produces population-based survival data for a entire geographical area covering 20% of the French population. Survival improves with earlier diagnosis but depends also on quality of care and availability of care: access to medical care can be a favouring factor. While in terms of incidence the situation in the Ile-de-France area is close to the national situation, survival at 5 years for a set of tumors appears to be better in this area than the nationwide figures. Besides providing information useful to determine mortality and incidence, the health insurance data offer additional insight to cancer epidemiology and contribute to better knowledge of this disease.

For pertinent investigation of patient satisfaction it is important to know what type of data collection procedure yields optimal results. In this study, we compared a self- versus interview administration of the European Organisation for Research and Treatment of Cancer (EORTC) patient satisfaction questionnaire (QLQ-SAT32), noting response and item omission rates. We also assessed the correlation and agreement between responses according to these modalities of questionnaire administration.123 patients recruited in two anti-cancer Centers in France were asked to complete the QLQ-SAT32 at home within 2 weeks of hospital discharge and were randomly allocated to participate in a telephone interview-administration of the QLQ-SAT32, either before or after self-completing the QLQ-SAT32.104 patients completed both modalities of questionnaire administration. Agreement proved excellent for the doctors' and nurses' subscales and satisfactory for the services' subscale and the general satisfaction single item. However, a significant difference in response rate showed up between Centers (different interviewers) and the interview modality revealed an item omission rate higher than for the self-administration modality.Self- and interview-administration of the QLQ-SAT32 provided similar responses however different response rate between interviewers and higher item omission rate in interviews suggests a biasing effect of that modality of data collection.

Because of long-term complications of external beam radiation in retinoblastoma, a new therapeutic modality using chemotherapy and local treatments is used whenever possible. We conducted a retrospective study to evaluate visual results.We studied visual acuity and ocular side effects in children presenting uni- or bilateral retinoblastoma in whom we were able to achieve conservative management without external beam treatment. The treatments that were used included chemotherapy, chemothermotherapy, diode laser, iodine 125 plaque brachytherapy, and cryotherapy. The initial characteristics (diameter, subretinal or vitreous seeding, and location) of each tumor, the treatments used, their results, and the visual function after age 4 years were recorded.We treated 429 children for retinoblastoma at the Curie Institute between October 1994 and December 2002. Two hundred twenty-seven eyes had conservative treatment without external beam. We were able to study the visual function in 102 eyes at a median age of 67 months. The median follow-up after the end of the treatment was 5 years. The mean visual acuity was 20/37 and 60% of the children had visual acuity of more than 20/40. Macular alteration was observed in 34 eyes. Statistical analysis showed that a larger tumor diameter of the retinal surface (p0.0003) and location close to the macula (p0.0001) were the most significant risk factors for vision loss.The visual results of the treatment of retinoblastoma by chemotherapy and local treatments is good when the tumors are not located close to the macula. Larger tumors also have a worse prognosis for vision.

Adjuvant radiotherapy to the breast plays a significant role in preventing local failure in women treated for early stage breast cancer. This fact is supported by multiple clinical trials demonstrating that adjuvant radiotherapy decreases the risk of local recurrence and increases the rate of breast preservation, and actually the rules of adjuvant breast irradiation are clearly established. Sarcomas are a rare but recognized complication of radiation therapy for breast carcinoma, and are associated with poor prognosis. The first case of a bone sarcoma after radiation therapy of breast cancer was described by Beck in 1922. In 1948, Cahan et al. defined the criteria for diagnosis of radiation-induced sarcoma. Since then, some studies have reported the incidence of radiation-induced sarcoma after radiotherapy for different cancers. This article reports and discusses the incidence, management and treatment outcome of radiation-induced sarcomas occurring after radiotherapy for breast cancer in our institute. The incidence, histology, latency of appearance, genesis, their treatment and the prognostic factors of these rare tumors are discussed and the literature is reviewed.

Central venous catheters are essential to management of cancer patients. Thrombotic complications are potentially severe, but it is difficult to prescribe systematically a prophylactic treatment. So it is necessary to identify the higher risk patients who need a prophylaxis.To identify factors associated with the development of clinically significant venous thrombosis in cancer patients with long-term catheters.Monocentric prospective study about 5447 long-term central venous catheters inserted into patients receiving treatment for solid tumours (50% of breast cancers). Clinically significant catheter-related thromboses are confirmed by ultrasonography, phlebography or scanner.The median duration of catheter use is 147 days. There are 135 clinically significant catheter-related thromboses. The risk for thrombosis is 0.1149 events per 1000 device days. The incidence of symptomatic thrombosis is 0.9% at 30 days, 1.36% at 60 days, 1.83% at 90 days and 2.25% at 120 days. The multivariate analysis shows that female sex, duration of insertion procedure (more than 25 minutes) and place of insertion (femoral place) are factors associated with clinically significant venous thrombosis. The right subclavian insertion causes less risk. The catheters with their tip too "high" in the superior vena cava are systematically changed, so that the position of the catheter tip does not appear as a risk factor in our study. The disease stage and the type of treatments are not well examined.A predictive score can be made with the three risk factors that have been identified. Thanks to this score it is possible to determine the patients with higher risk for clinically significant catheter-related venous thrombosis. These patients must be more frequently watched over and must receive a prophylactic treatment. The best prophylaxis has to be determined.

This retrospective study compared the rate of local recurrence after irradiation of uveal melanoma treated with iodine 125 plaques or proton beam therapy.Iodine 125 plaques were used to treat all uveal melanomas between the end of 1989 and 1991. Since 1991, we have used iodine plaques for small anterior tumors and proton beam for other tumors. We use a plaque with a larger diameter than the tumor diameter (2-4mm) with a dose of 90Gy at the apex. Proton beam therapy is used for all tumors at the equator or posterior to the equator not thicker than 12mm. The dose given is 60Gy cobalt equivalent in four fractions. For each patient, the initial size and location of the tumor were noted as well as the follow-up each year: the outcome for the eye (local recurrence, ocular conservation, and functional results), the occurrence of metastasis, and survival. A statistical analysis was performed.Between December 1989 and September 1998, 1272 patients were treated: 926 (72.8%) were treated with proton beam irradiation and 346 (27.8%) with iodine 125 plaques. The median follow-up was 5 years (60 months). For the patients treated with proton beam therapy, the mean age was 58 years, the tumor location was anterior to the equator for 3.8%, at the equator for 43.6%, and posterior to the equator for 52.6%. The mean tumor diameter was 13.4mm and the mean tumor thickness was 5.69mm. For the patients treated with iodine 125 plaques, the mean age was 61.5 years. The location of the tumor was anterior to the equator for 34.4%, at the equator for 46.5%, and posterior to the equator for 19.1%. The mean tumor diameter was 11.5mm and the mean tumor thickness was 5.12mm. The recurrence rate was 4% for the proton beam treatment and 3.75% for iodine plaques. There was no statistical difference.In the literature, the rate of local recurrence is usually higher with iodine 125 plaques than proton beam therapy. We discuss the risk factors for local recurrence after iodine 125 plaques: tumor diameter, lower dose to the tumor apex and lower dose rate, and posterior location of the tumor. We found a higher mortality rate in patients who presented local recurrence.When we use iodine 125 plaques for anterior tumors with the proper dose and dose rate to the apex of the tumor, we do not find more recurrence than with proton beam therapy.

Uveal malignant melanoma is the most common primary intraocular tumor in adults. The occurrence of bilateral uveal melanoma is an extremely rare event, but the observed frequency is nevertheless higher than what can be attributed to chance. Possible responsible factors may include a genetic predisposition.This retrospective study investigated the charts of patients examined from July 1988 to July 2001. For each patient, the clinical characteristics of the tumor (diameter, thickness, location), treatments, and results were noted, as were the eye involved, the presence of ocular melanocytosis, cutaneous melanoma, and second primary cancers. The information was then subjected to statistical analysis.Of 2 461 patients with unilateral primary uveal melanoma, five were identified as having bilateral uveal melanoma (0.2%). The expected number of cases would be less than one, hypothesizing an incidence of second melanoma identical to the incidence of a primary melanoma in the general population. The interval between the diagnosis of first and second primary uveal melanomas ranged from 0 to 6 years (median, 2 years). There was no clinical evidence of ocular melanocytosis in any of the five patients. The uveal melanoma was choroidal in three patients and affected the ciliary body or iris and choroid in two patients.The discrepancy between the estimated incidence (thought by Shammas to be one case every 18 years) and the observed incidence of bilateral primary uveal melanoma could be the result of many possible factors. An increased incidence of unilateral uveal melanoma could be a cause but in fact the incidence of uveal melanoma seems stable. Uveal melanoma may have been misdiagnosed in earlier years. The presence of a genetic predisposition to uveal melanoma is a possible explanation (suspected because of bilateral cases, familial cases and association with other primary malignancies). Ocular melanocytosis, which is described as more common in patients with bilateral uveal melanoma, was not seen in our series.Bilateral primary uveal melanoma occurs more frequently than expected. Unidentified germline mutations may be involved in pathogenesis. These cases serve as a reminder of the of the importance of careful examination of the second eye.

We have carried out a retrospective study of 135 patients followed at Curie Institute for choroidal naevus between March 1983 and June 1997.54 patients presented naevi considered as benign and 81 patients presented suspicious choroidal naevi (with at least one of the following findings: visual symptoms; serious detachment of the retina; orange pigment; thickness greater than 2mm or a diameter greater than 7mm). These suspicious naevi were followed more carefully. The median follow-up was 49 months. The median diameter of the lesions was 6mm and the median thickness was 1.5mm. We studied the age of the patients, clinical symptoms, the presence or absence of orange pigment, drusen and serious detachment, and the angiographic and echographic findings.3 patients died of unrelated cause; 7 patients were lost to follow-up; 30 patients presented documented growth; 4 of them belonged to the group considered as benign and 26 to the group considered as suspicious, with a significant difference between the groups. The lesions that grew were treated by proton beam or I125 patches. The risk factors for growth that were statistically significant were the presence of visual symptoms, pin points, orange pigment, serous detachment and thickness greater than 2mm. With drusen, the risk for growth was significantly less.These results are very similar to those published in the literature concerning risk factors for growth of choroidal nevi. The absence of metastatic spread in patients whose nevi have grown show that it is possible to monitor choroidal naevi if there is a doubt as to the malignant or benign nature of the lesion.It is important to determine if a choroidal naevus is suspicious or benign and to propose closer follow-up for suspicious lesions.

To report results of liver resections for breast cancer liver metastasis (BCLM) and to evaluate the rate of survival and the prognostic factors.Between 1988 and 1999, 69 patients were operated on for BCLM and 65 who had liver resection were analyzed. The selection criteria for surgery were: normal performance status and liver function test; radiological objective response to chemotherapy (and/or hormonotherapy); in cases of non-isolated BCLM, complete response of associated metastatic site (usually bone) and no brain metastases. The mean age of the 65 patients was 47 (30-70) years. BCLM was diagnosed an average of 60 (0-205) months after the initial cancer. The BCLM was more frequently solitary (n = 44). The mean diameter was 3.8 (0-12) cm. The mean number of cycles of chemotherapy before surgery was 7.5 (3-24). Liver resections included major hepatectomy (n = 31): right n = 19, extended left n = 4, left n = 8, minor hepatectomy (n = 25) and limited resection (n = 9).There was no postoperative mortality. The 18% morbidity rate included a majority of pleural effusions with two reoperations. The median follow-up was 41 months (6-100 months). The survival rate after surgery was 90% at 1 year, 71% at 3 and 46% at 4 years. Thirteen patients are alive at 4 years. The 36-month survival rate differed according to the time to onset of BCLM: 55% before versus 86% after 48 months (p = 0.01). The other studied factors were not statistically associated with survival. The recurrence rate in the remaining liver at 36 months differed according to the lymph node status of the initial breast cancer: 40% for N0-N1 versus 81% for N1b-N2 (p = 0.01) and according to the type of liver resection: 45% for minor liver resection versus 73% for major (p = 0.02).Adjuvant liver surgery should be included in multicenter treatment protocols for medically-controlled breast cancer liver metastasis.

In 1986, The Fédération Nationale desCentres de Lutte Contre le Cancer Breast Group initiated a multicenter randomized trial to assess the usefulness of long-term adjuvant tamoxifen treatment. Short-term adjuvant tamoxifen treatment was to be compared with life long adjuvant tamoxifen treatment.Patients who were disease-free after 2 to 3 years of adjuvant tamoxifen treatment were eligible for the trial. From September 1986 to May 1995, 3,793 patients were randomized from France, Belgium, and Argentina. A total of 1,882 patients stopped tamoxifen (short-term group), and 1,911 patients were to continue tamoxifen for life (long-term group) at the same dose as previously prescribed. The protocol was modified in February 1997, limiting tamoxifen treatment to 10 years after randomization, thus giving a comparison between a 2- to 3-year treatment and a 12- to 13-year treatment. To date, the median duration of tamoxifen treatment is 30 months in the short-term group, and 70 months in the long-term group.Overall, longer tamoxifen treatment induced a 23% reduction in relapse rates, leading to a 7-year disease-free survival rate of 78%, compared with 72% in the shorter-treatment group. In contrast, overall survival did not differ between the two groups, with a 79% overall survival rate in both groups. This improvement in disease-free survival could be observed in node-positive patients (P: =.001); however, it was not found in node-negative patients. Prolonged tamoxifen treatment corresponded to a significant increase in disease-free survival in estrogen receptor-positive patients (P: =.03) as well as in estrogen receptor-negative patients (P: =.05). Furthermore, longer treatment reduced contralateral breast cancers and did not increase the number of endometrial cancers.Although no survival advantage was noted, patients did benefit from longer tamoxifen treatment over 3 years and had significantly better disease-free survival compared with patients who stopped hormonal treatment. Long-term follow-up is needed to assess these results. Most patients in the long-term group are still receiving treatment. Comparison of results as time passes will enable conclusions to be made on the value of long-term treatment over 5 years compared with 2 to 3 years.